diff options
| author | Eric Dumazet <edumazet@google.com> | 2024-03-28 14:40:32 +0000 |
|---|---|---|
| committer | Jakub Kicinski <kuba@kernel.org> | 2024-03-29 15:03:11 -0700 |
| commit | 81e3f608479f6febf81d3ac84fc985cf45165764 (patch) | |
| tree | 3911d590615aae3fe7a87fc61726b6650670b822 | |
| parent | 5d02f9af7a1ffc7d3f11da77e2cd9cb878d403c9 (diff) | |
| download | linux-crypto-81e3f608479f6febf81d3ac84fc985cf45165764.tar.gz linux-crypto-81e3f608479f6febf81d3ac84fc985cf45165764.zip | |
net: add sk_wake_async_rcu() helper
While looking at UDP receive performance, I saw sk_wake_async()
was no longer inlined.
This matters at least on AMD Zen1-4 platforms (see SRSO)
This might be because rcu_read_lock() and rcu_read_unlock()
are no longer nops in recent kernels ?
Add sk_wake_async_rcu() variant, which must be called from
contexts already holding rcu lock.
As SOCK_FASYNC is deprecated in modern days, use unlikely()
to give a hint to the compiler.
sk_wake_async_rcu() is properly inlined from
__udp_enqueue_schedule_skb() and sock_def_readable().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20240328144032.1864988-5-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
| -rw-r--r-- | crypto/af_alg.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/crypto/af_alg.c b/crypto/af_alg.c index 68cc9290..5bc6d0fa 100644 --- a/crypto/af_alg.c +++ b/crypto/af_alg.c @@ -847,7 +847,7 @@ void af_alg_wmem_wakeup(struct sock *sk) wake_up_interruptible_sync_poll(&wq->wait, EPOLLIN | EPOLLRDNORM | EPOLLRDBAND); - sk_wake_async(sk, SOCK_WAKE_WAITD, POLL_IN); + sk_wake_async_rcu(sk, SOCK_WAKE_WAITD, POLL_IN); rcu_read_unlock(); } EXPORT_SYMBOL_GPL(af_alg_wmem_wakeup); @@ -914,7 +914,7 @@ static void af_alg_data_wakeup(struct sock *sk) wake_up_interruptible_sync_poll(&wq->wait, EPOLLOUT | EPOLLRDNORM | EPOLLRDBAND); - sk_wake_async(sk, SOCK_WAKE_SPACE, POLL_OUT); + sk_wake_async_rcu(sk, SOCK_WAKE_SPACE, POLL_OUT); rcu_read_unlock(); } |